Home » Programme » Workshops » Workshop 3: Software Development for Scientific Applications: current and future perspectives


Workshop 3: Software Development for Scientific Applications: current and future perspectives

Organisers

James Annett, Bruce Beckles, Chris Greenough, Neil Chue Hong, Peter Kilpatrick, Stan Scott

The following two complementary workshops aim to foster a symbiotic relationship amongst the computer science, computational science and software engineering communities - members of the CCP community are particularly encouraged to participate in both workshops. Participants will explore current best practice and future trends in software engineering and development techniques and their use in the construction of quality scientific software applications.

Workshop A: Scientific Software development in the UK, the CCP model

Workshop B: Reliable and Efficient Computational Science Software in Dynamic Grid Environments

Workshop A will take place in the morning and Workshop B in the afternoon. Participants are encouraged to attend both workshops. There will be a combined round table discussion immediately following Workshop B involving participants from both workshops.

Workshop A: Scientific Software development in the UK, the CCP model

Workshop Organiser(s): James Annett

The collaborative Computational Projects are the UK 's unique approach to developing and maintaining high quality scientific software, and making it widely available to UK academic users. They are a group of collaborations which bring together leading UK expertise in key fields of computational research. They tackle large scale scientific software development, maintenance, and distribution and support of high quality codes. Each project represents many years of intellectual and financial investment in a given scientific area. The areas of science which are supported by CCP projects in the UK include, Quantum Chemistry, Atomic and Molecular Physics, Surface Science, Protein Crystallography, Molecular Simulation, Molecular Quantum Dynamics, Electronic Structure of Solids, Biosequences and Function, Computational Engineering, Fibre Diffraction, Powder and Single Crystal Diffraction, NMR Spectroscopy, Biomolecular Simulation, and more projects are continually being added to this list.

The aim of the AHM2008 workshop is to bring together leading scientists and software developers from the CCPs to present recent scientific results and also to compare and share advances in software development methodologies and strategies. These can include use of the CCPForge code development site, XML formats and emerging standards for each scientific area, Visualization tools and GUIs developed for CCP software, High Performance Computing and parallelization towards the petaflop scale, GNU and other license models for code releases and other issues where the CCP and e-science communities can benefit from shared experience and collaboration.

Workshop B: Reliable and efficient computational science software in dynamic grid environments

Organiser(s):

Bruce Beckles

University of Cambridge Computing Service

Chris Greenough

Rutherford Appleton Laboratory

Neil Chue Hong

OMII-UK

Peter Kilpatrick

Queen's University of Belfast

Stan Scott

Queen's University of Belfast

For more information contact: Stan Scott at ns.scott@ qub.ac.uk

Description

The goal of this workshop is to provide a forum for the exchange of ideas on modern and innovative software engineering methodologies and tools that will enable the development of reliable and efficient computational science software in distributed and high performance computing environments.

The exploitation of grid infrastructures by computational scientists places significant additional demands on application software developers. For example, in a dynamic grid environment computational software needs the ability to adapt itself at run time to cope with changing user needs, changing operational environments, system faults and resource variability. Accordingly, the software should, within quality of service constraints, be able to configure and reconfigure itself as the need arises and to hide its complexity from the user.

The workshop aims to foster a symbiotic relationship amongst the computer science, computational science and software engineering communities and enable participants to exchange and debate future trends in software engineering and their application to computational science. Participants will have practical and/or research experience in:

  • developing computational science applications software; and/or
  • developing, extending, deploying or using software engineering methodologies and tools.

Members of the Collaborative Computational Projects (CCPs) are particularly encouraged to attend.

Call for Papers

The workshop will take the form of a symposium with one invited keynote talk and several submitted talks . There will be a combined round table discussion immediately following this workshop involving participants from both workshops.

We will invite contributions around the following broad themes:

  • Reliability – including numerical validation and fault-tolerance.
  • Efficiency – including algorithms and load balancing.
  • Adaptivity – including changing operational environments and resource variability.
  • Software Engineering Tools
  • Software Engineering Methodologies – including agile, monumental, distributed, component-based technologies, skeleton systems and structured parallel programming.
  • Requirements Engineering
  • Usability Engineering
  • Community - including community development, community engagement and user driven development.



Home | About | Contact Us | Registration
e-Science InstituteUniversity of EdinburghSchool of Informatics, University of EdinburghNational e-Science Centre
Photographs by P Tuffy